indri::api::IndexEnvironment Class Reference

Principal class for interacting with Indri indexes during index construction. Provides the API for opening or creating an index and its associated repository, setting indexing and text parsing parameters, and adding documents to the repository. More...

#include <IndexEnvironment.hpp>

List of all members.

Public Member Functions
	IndexEnvironment ()
	~IndexEnvironment ()
void	setOffsetAnnotationsPath (const std::string &offsetAnnotationsRoot)
void	setOffsetMetadataPath (const std::string &offsetMetadataRoot)
void	setAnchorTextPath (const std::string &anchorTextRoot)
void	setDocumentRoot (const std::string &documentRoot)
void	addFileClass (const std::string &name, const std::string &iterator, const std::string &parser, const std::string &tokenizer, const std::string &startDocTag, const std::string &endDocTag, const std::string &endMetadataTag, const std::vector< std::string > &include, const std::vector< std::string > &exclude, const std::vector< std::string > &index, const std::vector< std::string > &metadata, const std::map< indri::parse::ConflationPattern *, std::string > &conflations)
indri::parse::FileClassEnvironmentFactory::Specification *	getFileClassSpec (const std::string &name)
void	addFileClass (const indri::parse::FileClassEnvironmentFactory::Specification &spec)
void	setIndexedFields (const std::vector< std::string > &fieldNames)
void	setNumericField (const std::string &fieldName, bool isNumeric, const std::string &parserName="")
void	setOrdinalField (const std::string &fieldName, bool isOrdinal)
void	setParentalField (const std::string &fieldName, bool isParental)
void	setMetadataIndexedFields (const std::vector< std::string > &forwardFieldNames, const std::vector< std::string > &backwardFieldNames)
void	setStopwords (const std::vector< std::string > &stopwords)
void	setStemmer (const std::string &stemmer)
void	setMemory (UINT64 memory)
void	setNormalization (bool flag)
void	setStoreDocs (bool flag)
void	setOffsetAnnotationIndexHint (indri::parse::OffsetAnnotationIndexHint hintType)
void	create (const std::string &repositoryPath, IndexStatus *callback=0)
void	open (const std::string &repositoryPath, IndexStatus *callback=0)
void	close ()
	close the index and repository
void	addFile (const std::string &fileName)
void	addFile (const std::string &fileName, const std::string &fileClass)
lemur::api::DOCID_T	addString (const std::string &documentString, const std::string &fileClass, const std::vector< indri::parse::MetadataPair > &metadata)
lemur::api::DOCID_T	addString (const std::string &documentString, const std::string &fileClass, const std::vector< indri::parse::MetadataPair > &metadata, const std::vector< indri::parse::TagExtent * > &tags)
lemur::api::DOCID_T	addParsedDocument (ParsedDocument *document)
void	deleteDocument (lemur::api::DOCID_T documentID)
int	documentsIndexed ()
	Returns the number of documents indexed so far in this session.
int	documentsSeen ()
void	compact ()
Static Public Member Functions
void	merge (const std::string &outputIndex, const std::vector< std::string > &inputIndexes)
Private Member Functions
void	_getParsingContext (indri::parse::Parser **parser, indri::parse::Tokenizer **tokenizer, indri::parse::DocumentIterator **iterator, indri::parse::Conflater **conflater, const std::string &extension)
std::vector< indri::parse::Transformation * >	_createAnnotators (const std::string &fileName, const std::string &fileClass, indri::parse::Conflater **conflater)
ParsedDocument *	_applyAnnotators (std::vector< indri::parse::Transformation * > &annotators, ParsedDocument *parsed)
Private Attributes
IndexStatus *	_callback
Parameters *	_options
std::string	_repositoryPath
indri::collection::Repository	_repository
int	_documents
std::string	_error
std::string	_offsetAnnotationsRoot
std::string	_offsetMetadataRoot
std::string	_anchorTextRoot
std::string	_documentRoot
Parameters	_parameters
indri::parse::FileClassEnvironmentFactory	_fileClassFactory
indri::parse::AnchorTextAnnotator	_annotator
indri::parse::OffsetAnnotationAnnotator	_oa_annotator
indri::parse::OffsetMetadataAnnotator	_om_annotator
std::map< std::string, indri::parse::FileClassEnvironment * >	_environments
int	_documentsIndexed
int	_documentsSeen
Friends
class	QueryEnvironment

---

Detailed Description

Principal class for interacting with Indri indexes during index construction. Provides the API for opening or creating an index and its associated repository, setting indexing and text parsing parameters, and adding documents to the repository.

---

Constructor & Destructor Documentation

indri::api::IndexEnvironment::IndexEnvironment (   )

indri::api::IndexEnvironment::~IndexEnvironment (   )

---

Member Function Documentation

indri::api::ParsedDocument * indri::api::IndexEnvironment::_applyAnnotators (  std::vector< indri::parse::Transformation * > &  annotators, ParsedDocument *  parsed )  [private]

std::vector< indri::parse::Transformation * > indri::api::IndexEnvironment::_createAnnotators (  const std::string &  fileName, const std::string &  fileClass, indri::parse::Conflater **  conflater )  [private]

void indri::api::IndexEnvironment::_getParsingContext (  indri::parse::Parser **  parser, indri::parse::Tokenizer **  tokenizer, indri::parse::DocumentIterator **  iterator, indri::parse::Conflater **  conflater, const std::string &  extension )  [private]

void indri::api::IndexEnvironment::addFile (  const std::string &  fileName, const std::string &  fileClass )

add a file of the specified file class to the index and repository Parameters: fileName  the file to add fileClass  the file class to add (eg trecweb).

void indri::api::IndexEnvironment::addFile (  const std::string &  fileName  )

Add the text in a file to the index and repository. The fileClass of this file will be chosen based on the file extension. If the file has no extension, it will be skipped. Information about indexing progress will be passed to the callback. See also: setCallback() Parameters: fileName  the file to add

void indri::api::IndexEnvironment::addFileClass (  const indri::parse::FileClassEnvironmentFactory::Specification &  spec  )  [inline]

Add a file class. Parameters: spec  The file class to add.

void indri::api::IndexEnvironment::addFileClass (  const std::string &  name, const std::string &  iterator, const std::string &  parser, const std::string &  tokenizer, const std::string &  startDocTag, const std::string &  endDocTag, const std::string &  endMetadataTag, const std::vector< std::string > &  include, const std::vector< std::string > &  exclude, const std::vector< std::string > &  index, const std::vector< std::string > &  metadata, const std::map< indri::parse::ConflationPattern *, std::string > &  conflations )

Add parsing information for a file class. Data for these parameters is passed into the FileClassEnvironmentFactory Parameters: name  name of this file class, eg trecweb iterator  document iterator for this file class parser  document parser for this file class tokenizer  document tokenizer for this file class startDocTag  tag indicating start of a document endDocTag  tag indicating the end of a document endMetadataTag  tag indicating the end of the metadata fields include  default tags whose contents should be included in the index exclude  tags whose contents should be excluded from the index index  tags that should be forwarded to the index for tag extents metadata  tags whose contents should be indexed as metadata conflations  tags that should be conflated

lemur::api::DOCID_T indri::api::IndexEnvironment::addParsedDocument (  ParsedDocument *  document  )

add an already parsed document to the index and repository Parameters: document  the document to add

lemur::api::DOCID_T indri::api::IndexEnvironment::addString (  const std::string &  documentString, const std::string &  fileClass, const std::vector< indri::parse::MetadataPair > &  metadata, const std::vector< indri::parse::TagExtent * > &  tags )

lemur::api::DOCID_T indri::api::IndexEnvironment::addString (  const std::string &  documentString, const std::string &  fileClass, const std::vector< indri::parse::MetadataPair > &  metadata )

Adds a string to the index and repository. The documentString is assumed to contain the kind of text that would be found in a file of type fileClass. Parameters: documentString  the document to add fileClass  the file class to add (eg trecweb). metadata  the metadata pairs associated with the string.

void indri::api::IndexEnvironment::close (   )

close the index and repository

void indri::api::IndexEnvironment::compact (   )

Permanently deletes information for documents that have been deleted from the index and reclaims used disk space.

void indri::api::IndexEnvironment::create (  const std::string &  repositoryPath, IndexStatus *  callback = 0 )

create a new index and repository Parameters: repositoryPath  the path to the repository callback  IndexStatus object to be notified of indexing progress.

void indri::api::IndexEnvironment::deleteDocument (  lemur::api::DOCID_T  documentID  )

Delete an existing document. Parameters: documentID  The document to delete.

int indri::api::IndexEnvironment::documentsIndexed (   )

Returns the number of documents indexed so far in this session.

int indri::api::IndexEnvironment::documentsSeen (   )

Returns the number of documents considered for indexing, which is the sum of the documents indexed and the documents skipped.

indri::parse::FileClassEnvironmentFactory::Specification* indri::api::IndexEnvironment::getFileClassSpec (  const std::string &  name  )  [inline]

Get a named file class. Parameters: name  The name of the file class to retrieve.

void indri::api::IndexEnvironment::merge (  const std::string &  outputIndex, const std::vector< std::string > &  inputIndexes )  [static]

Merges the contents of the indexes referenced in the inputIndexes list and creates a new index called outputIndex. The final index is compacted (contains no information about deleted documents). Parameters: outputIndex  The pathname to the index to create (should not exist yet). inputIndexes  The pathnames to indexes to merge. These indexes should not currently be open.

void indri::api::IndexEnvironment::open (  const std::string &  repositoryPath, IndexStatus *  callback = 0 )

open an existing index and repository Parameters: repositoryPath  the path to the repository callback  IndexStatus object to be notified of indexing progress.

void indri::api::IndexEnvironment::setAnchorTextPath (  const std::string &  anchorTextRoot  )

Set anchor text root path. Parameters: anchorTextRoot  path to anchor text root.

void indri::api::IndexEnvironment::setDocumentRoot (  const std::string &  documentRoot  )

Set the document root path Parameters: documentRoot  path to document root.

void indri::api::IndexEnvironment::setIndexedFields (  const std::vector< std::string > &  fieldNames  )

Set names of fields to be indexed. This call indicates to the index that information about these fields should be stored in the index so they can be used in queries. This does not affect whether or not the text in a particular field is stored in an index. See also: addFileClass Parameters: fieldNames  the list of fields.

void indri::api::IndexEnvironment::setMemory (  UINT64  memory  )

set the amount of memory to use for internal structures Parameters: memory  the number of bytes to use.

void indri::api::IndexEnvironment::setMetadataIndexedFields (  const std::vector< std::string > &  forwardFieldNames, const std::vector< std::string > &  backwardFieldNames )

Set names of metadata fields to be indexed for fast retrieval. The forward fields are indexed in a B-Tree mapping (documentID, metadataValue). If a field is not forward indexed, the documentMetadata calls will still work, but they will be slower (the document has to be retrieved, decompressed and parsed to get the metadata back, instead of just a B-Tree lookup). The backward indexed fields store a mapping of (metadataValue, documentID). If a field is not backward indexed, the documentIDsFromMetadata and documentFromMetadata calls will not work. Parameters: forwardFieldNames  the list of fields to forward index. backwardFieldNames  the list of fields to backward index.

void indri::api::IndexEnvironment::setNormalization (  bool  flag  )

set normalization of case and some punctuation; default is true (normalize during indexing and at query time) Parameters: flag  True, if text should be normalized, false otherwise.

void indri::api::IndexEnvironment::setNumericField (  const std::string &  fieldName, bool  isNumeric, const std::string &  parserName = "" )

Set the numeric property of a field. Parameters: fieldName  the field. isNumeric  true if the field is a numeric field, false if not. parserName  The name of the Transformation to use to compute the numeric value of the field. Repository currently recognizes the name NumericFieldAnnotator.

void indri::api::IndexEnvironment::setOffsetAnnotationIndexHint (  indri::parse::OffsetAnnotationIndexHint  hintType  )

provides the indexer with the hint strategy to use for speed optimizations for indexing offset annotations Parameters: hintType  the int type (of OffsetAnnotationIndexHint enum type)

void indri::api::IndexEnvironment::setOffsetAnnotationsPath (  const std::string &  offsetAnnotationsRoot  )

Set offset annotations root path. Parameters: offsetAnnotationsRoot  path to offset annotations root.

void indri::api::IndexEnvironment::setOffsetMetadataPath (  const std::string &  offsetMetadataRoot  )

Set offset metadata root path. Parameters: offsetMetadataRoot  path to offset metadata root.

void indri::api::IndexEnvironment::setOrdinalField (  const std::string &  fieldName, bool  isOrdinal )

Set the ordinal property of a field. Parameters: fieldName  the field. isOrdinal  true if the field is an ordinal field, false if not.

void indri::api::IndexEnvironment::setParentalField (  const std::string &  fieldName, bool  isParental )

Set the parental property of a field. Parameters: fieldName  the field. isParental  true if the field stores its parent, false if not

void indri::api::IndexEnvironment::setStemmer (  const std::string &  stemmer  )

set the stemmer to use Parameters: stemmer  the stemmer to use. One of krovetz, porter

void indri::api::IndexEnvironment::setStopwords (  const std::vector< std::string > &  stopwords  )

set the list of stopwords Parameters: stopwords  the list of stopwords

void indri::api::IndexEnvironment::setStoreDocs (  bool  flag  )

set storing of ParsedDocuments; default is true Parameters: flag  true, if ParsedDocuments should be stored, false otherwise.

---

Friends And Related Function Documentation

friend class QueryEnvironment [friend]

---

Member Data Documentation

std::string indri::api::IndexEnvironment::_anchorTextRoot [private]

indri::parse::AnchorTextAnnotator indri::api::IndexEnvironment::_annotator [private]

IndexStatus* indri::api::IndexEnvironment::_callback [private]

std::string indri::api::IndexEnvironment::_documentRoot [private]

int indri::api::IndexEnvironment::_documents [private]

int indri::api::IndexEnvironment::_documentsIndexed [private]

int indri::api::IndexEnvironment::_documentsSeen [private]

std::map<std::string, indri::parse::FileClassEnvironment*> indri::api::IndexEnvironment::_environments [private]

std::string indri::api::IndexEnvironment::_error [private]

indri::parse::FileClassEnvironmentFactory indri::api::IndexEnvironment::_fileClassFactory [private]

indri::parse::OffsetAnnotationAnnotator indri::api::IndexEnvironment::_oa_annotator [private]

std::string indri::api::IndexEnvironment::_offsetAnnotationsRoot [private]

std::string indri::api::IndexEnvironment::_offsetMetadataRoot [private]

indri::parse::OffsetMetadataAnnotator indri::api::IndexEnvironment::_om_annotator [private]

Parameters* indri::api::IndexEnvironment::_options [private]

Parameters indri::api::IndexEnvironment::_parameters [private]

indri::collection::Repository indri::api::IndexEnvironment::_repository [private]

std::string indri::api::IndexEnvironment::_repositoryPath [private]

---

The documentation for this class was generated from the following files:

• IndexEnvironment.hpp
• IndexEnvironment.cpp

---