SEXP generateResults(string _qno, string _query, int _documentLimit, bool stats) {
resultsData = resultsData_nullCopy;
documentIDs.clear();
scores.clear();
extDocIDs.clear();
terms.clear();
_gramTable.clear();
results.clear();
qno = _qno;
query = _query;
documentLimit = _documentLimit;
qa = environment.runAnnotatedQuery(query, _documentLimit);
results = qa->getResults();
_logtoposterior(results);
// Extract Documents
for (size_t i = 0; i < results.size(); i++){
documentIDs.push_back(results[i].document);
scores.push_back(results[i].score);
}
extDocIDs = environment.documentMetadata(documentIDs, "docno");
if(stats){
updateQueryDetails(environment, resultsData, query);
countGrams();
buildStats();
}
return Rcpp::wrap(true);
}
SEXP runQuery(string _qno, string _query, int _documentLimit, string _runid="default"){
indri::api::QueryAnnotation* qa;
qa = environment.runAnnotatedQuery(_query, _documentLimit);
std::vector<indri::api::ScoredExtentResult> results = qa->getResults();
//_logtoposterior(results);
// Extract Documents
std::vector<lemur::api::DOCID_T> documentIDs;
std::vector<double> scores;
for (size_t i = 0; i < results.size(); i++){
documentIDs.push_back(results[i].document);
scores.push_back(results[i].score);
}
vector<string> res_qno;
vector<string> res_q0;
vector<string> res_runid;
int documentLimit = _documentLimit;
for(int i=0; i < documentLimit; i++){
res_qno.push_back(qno);
res_q0.push_back("Q0");
res_runid.push_back(_runid);
}
std::vector<string> extDocIDs = environment.documentMetadata(documentIDs, "docno");
return Rcpp::DataFrame::create( Named("topic")= _qno,
Named("q0")= res_q0, Named("docID")= wrap(extDocIDs),
Named("rank")= seq( 1, documentLimit ),
Named("score")= wrap(scores),
Named("runID")= res_runid);
}
void _printResultRegion( std::stringstream& output, std::string queryIndex, int start, int end ) {
std::vector<std::string> documentNames;
std::vector<indri::api::ParsedDocument*> documents;
std::vector<indri::api::ScoredExtentResult> resultSubset;
resultSubset.assign( _results.begin() + start, _results.begin() + end );
// Fetch document data for printing
if( _printDocuments || _printPassages || _printSnippets ) {
// Need document text, so we'll fetch the whole document
documents = _environment.documents( resultSubset );
documentNames.clear();
for( size_t i=0; i<resultSubset.size(); i++ ) {
indri::api::ParsedDocument* doc = documents[i];
std::string documentName;
indri::utility::greedy_vector<indri::parse::MetadataPair>::iterator iter = std::find_if( documents[i]->metadata.begin(),
documents[i]->metadata.end(),
indri::parse::MetadataPair::key_equal( "docno" ) );
if( iter != documents[i]->metadata.end() )
documentName = (char*) iter->value;
// store the document name in a separate vector so later code can find it
documentNames.push_back( documentName );
}
} else {
// We only want document names, so the documentMetadata call may be faster
documentNames = _environment.documentMetadata( resultSubset, "docno" );
}
std::vector<std::string> pathNames;
if ( _inexFormat ) {
// retrieve path names
pathNames = _environment.pathNames( resultSubset );
}
// Print results
for( size_t i=0; i < resultSubset.size(); i++ ) {
int rank = start+i+1;
std::string queryNumber = queryIndex;
if( _trecFormat ) {
// TREC formatted output: queryNumber, Q0, documentName, rank, score, runID
output << queryNumber << " "
<< "Q0 "
<< documentNames[i] << " "
<< rank << " "
<< resultSubset[ i ].score << " "
<< _runID << std::endl;
} else if( _inexFormat ) {
output << " <result>" << std::endl
<< " <file>" << documentNames[i] << "</file>" << std::endl
<< " <path>" << pathNames[i] << "</path>" << std::endl
<< " <rsv>" << resultSubset[i].score << "</rsv>" << std::endl
<< " </result>" << std::endl;
}
else {
// score, documentName, firstWord, lastWord
output << resultSubset[i].score << "\t"
<< documentNames[i] << "\t"
<< resultSubset[i].begin << "\t"
<< resultSubset[i].end << std::endl;
}
if( _printDocuments ) {
output << documents[i]->text << std::endl;
}
if( _printPassages ) {
int byteBegin = documents[i]->positions[ resultSubset[i].begin ].begin;
int byteEnd = documents[i]->positions[ resultSubset[i].end-1 ].end;
output.write( documents[i]->text + byteBegin, byteEnd - byteBegin );
output << std::endl;
}
if( _printSnippets ) {
indri::api::SnippetBuilder builder(false);
output << builder.build( resultSubset[i].document, documents[i], _annotation ) << std::endl;
}
if( documents.size() )
delete documents[i];
}
}
SEXP getMetaData(string metaDataKey){
vector<string> metaDataString = environment.documentMetadata(documentIDs, metaDataKey);
CharacterVector c = wrap(metaDataString);
c.attr("names") = extDocIDs;
return c;
}