int main(int argc, char **argv)
{
// Simplest possible options parsing: we just require two or more
// parameters.
if (argc < 3) {
cout << "usage: " << argv[0] << " <path to database> <search terms>" << endl;
exit(1);
}
// Catch any Xapian::Error exceptions thrown
try {
// Make the database
Xapian::Database db(argv[1]);
// Start an enquire session
Xapian::Enquire enquire(db);
// Set percent and/or weight cutoffs
enquire.set_cutoff(90,0.2);
// Set weighting schema
BM25Weight bm1(1.0,0.0,1.0,0.5,0.3);
enquire.set_weighting_scheme(bm1);
// Build the query object
Xapian::Query query(Xapian::Query::OP_AND, argv + 2, argv + argc);
cout << "Performing query" << query.get_description() << "'" << endl;
// Set Stopper
string stop[8]={"的","了","呵","吧","就","你","我","他"};
SimpleStopper *ss=new SimpleStopper;
for(int i=0;i<8;i++){
ss->add(stop[i]);
}
QueryParser qparser;
qparser.set_stopper(ss);
qparser.set_database(db);
// Give the query object to the enquire session
enquire.set_query(query);
// Get the top 10 results of the query
Xapian::MSet matches = enquire.get_mset(0, 10); //最多返回10个文档
// Display the results
cout << matches.size() << " results found" << endl;
for (Xapian::MSetIterator i = matches.begin();i != matches.end(); ++i) {
Xapian::Document doc = i.get_document();
cout << "Document ID " << *i << "\nPercent " <<i.get_percent() << "%\n" << doc.get_data() << "\n" << endl;
}
db.close();
} catch(const Xapian::Error &error) {
cout << "Exception: " << error.get_msg() << endl;
}
}
int
get(int argc, char** argv) {
// parse arguments
parseArguments(argc, argv);
string backend = options['t'];
string indexdir = options['d'];
// check arguments: indexdir
if (indexdir.length() == 0) {
pe("Provide the directory with the index.\n");
return usage(argc, argv);
}
// check arguments: dirs
if (arguments.size() == 0) {
pe("'%s' '%s'\n", backend.c_str(), indexdir.c_str());
pe("Provide one or more files to search.\n");
return usage(argc, argv);
}
// create an index manager
IndexManager* manager = getIndexManager(backend, indexdir);
if (manager == 0) {
return usage(argc, argv);
}
IndexReader* reader = manager->indexReader();
QueryParser parser;
for (vector<string>::iterator iter = arguments.begin();
iter != arguments.end(); ++iter) {
Query query = parser.buildQuery( "system.location:'"+ *iter + '\'');
vector<IndexedDocument> matches = reader->query(query, 0, 10);
if (matches.size() == 0)
printf ("%s: is not indexed\n", iter->c_str());
else
{
printf ("Information associated to %s:\n", iter->c_str());
for (vector<IndexedDocument>::iterator it = matches.begin();
it != matches.end(); ++it)
{
printIndexedDocument(*it);
}
}
}
IndexPluginLoader::deleteIndexManager(manager);
return 0;
}
// static
Query* MultiFieldQueryParser::parse(const TCHAR* query, const TCHAR** _fields, const uint8_t* flags, Analyzer* analyzer) {
BooleanQuery* bQuery = _CLNEW BooleanQuery();
for (size_t i = 0; _fields[i]!=NULL; i++) {
//TODO: this is really confusing... why not refactor _fields and flags to use a array object.
//flags can be NULL since NULL == 0...
/*if (flags[i] == NULL) {
_CLLDELETE(bQuery);
_CLTHROWA(CL_ERR_IllegalArgument, "_fields.length != flags.length");
}*/
QueryParser* qp = _CLNEW QueryParser(_fields[i], analyzer);
Query* q = qp->parse(query);
if (q!=NULL && // q never null, just being defensive
(!(q->instanceOf(BooleanQuery::getClassName())) || ((BooleanQuery*)q)->getClauseCount()>0)) {
bQuery->add(q, true, (BooleanClause::Occur)flags[i]);
} else
_CLLDELETE(q);
_CLLDELETE(qp);
}
return bQuery;
}
//static
Query* MultiFieldQueryParser::parse(const TCHAR** _queries, const TCHAR** _fields, Analyzer* analyzer)
{
BooleanQuery* bQuery = _CLNEW BooleanQuery();
for (size_t i = 0; _fields[i]!=NULL; i++)
{
if (_queries[i] == NULL) {
_CLLDELETE(bQuery);
_CLTHROWA(CL_ERR_IllegalArgument, "_queries.length != _fields.length");
}
// TODO: Reuse qp instead of creating it over and over again
QueryParser* qp = _CLNEW QueryParser(_fields[i], analyzer);
Query* q = qp->parse(_queries[i]);
if (q!=NULL && // q never null, just being defensive
(!(q->instanceOf(BooleanQuery::getClassName()) || ((BooleanQuery*)q)->getClauseCount() > 0))) {
bQuery->add(q, true, BooleanClause::SHOULD);
} else
_CLLDELETE(q);
_CLLDELETE(qp);
}
return bQuery;
}
int main(int argc, char* argv[])
{
// Get the top level suite from the registry
CPPUNIT_NS::Test *suite = CPPUNIT_NS::TestFactoryRegistry::getRegistry().makeTest();
CppUnit::TextUi::TestRunner runner;
//parse the SIMPLE soource code into the parser
//Parse();
string t;
Parser* parser;
PKBController* controller;
parser = new Parser("../Debug/Sample-Source.txt");
controller = parser->controller;
//string query = "assign a; Select a pattern a(\"oSCar\", \"1 + beta + tmp\" )";
string query = "assign a; Select a pattern a(\"oSCar\", \"_\" )";
QueryParser* qp = new QueryParser(query);
QueryEvaluator* qe = new QueryEvaluator(controller);
qp->getQueryTree()->printTree();
cout << qp->getQueryTree()->getRootNode()->getChild(2)->getKey() << endl;;
cin >> t;
//Call of DesignExtractor
//Extract();
runner.addTest(suite);
bool wasSucessful = runner.run();
//getchar();
return wasSucessful ? 0 : 1;
}
EngineQuery SearchStore::constructContainsQuery(const QByteArray& prefix, const QString& value)
{
QueryParser parser;
return parser.parseQuery(value, prefix);
}
void queryDriver(string query, list<string> &result, PKB *pkb){
//// cout<<"Begin parse query"<<endl;
pkb->printParentTable();
pkb->printModifiesTable();
pkb->printModifiesProcTable();
/*cout<<"test here"<<endl;
cout<<"getChildren 15"<<endl;
set<STMTNUM> temp = pkb->getChildren(15);
for(set<STMTNUM>::iterator it = temp.begin();it!=temp.end();it++){
cout<<*it<<endl;
}
cout<<"getParent 63"<<endl;
cout<<pkb->getParent(63)<<endl;*/
QueryParser qp;
QueryEvaluator qe = QueryEvaluator::QueryEvaluator(pkb);
bool isValid = true;
Query parsedQuery = qp.queryParse(query,isValid);
if(!isValid){
//// cout << "Query Invalid" << endl;
return;
}
//// cout<<"End parse query"<<endl;
//// cout<<"Begin evaluate query"<<endl;
vector<Relationship> newRelations;
unordered_map<string, int> synIndexMap;
unordered_map<int, vector<Pair>> clauseAnswers = qe.evaluateQuery(parsedQuery, &newRelations);
//// cout<<"End evaluate query"<<endl;
//// cout <<"Begin projecting results"<< endl;
vector<vector<int>> tupleTable = createTupleTable(clauseAnswers, newRelations, &synIndexMap );
vector<string> selectedSyn = parsedQuery.getSelectedSyn();
unordered_map<string, TypeTable::SynType> synTable = parsedQuery.getSynTable();
set<string> ansSet;
ansSet.insert("");
if(synTable.find(selectedSyn.at(0))->second == TypeTable::BOOLEAN) {
if(tupleTable.size()!=0)
result.push_back("true");
else
result.push_back("false");
return;
}
//// cout << "Begin handling unprocessed selected synonyms" << endl;
for(int i=0; i<int(selectedSyn.size()); i++) {
//// cout << "Checking for the following syn: " << selectedSyn.at(i) << endl;
if(synIndexMap.count(selectedSyn.at(i)) == 0) {
//// cout << selectedSyn.at(i) << " has not been processed" << endl;
unordered_map<string, TypeTable::SynType>::iterator it = synTable.find(selectedSyn.at(i));
vector<vector<int>> newTable;
set<int> allOfType;
if(it->second == TypeTable::VARIABLE) {
allOfType = pkb->getAllVarIndex();
}
else if(it->second == TypeTable::CONSTANT) {
set<int> allConstIndex = pkb->getAllConstIndex();
ConstTable* constTable = pkb->getConstTable();
for(set<int>::iterator it=allConstIndex.begin(); it!=allConstIndex.end(); it++) {
allOfType.insert(atoi(constTable->getConst(*it).c_str()));
}
}
else if(it->second == TypeTable::PROCEDURE) {
allOfType = pkb->getAllProcIndexes();
}
else if(it->second == TypeTable::STMTLST) {
allOfType = pkb->getAllStmtList();
}
else {
allOfType = pkb->getAllStmts(it->second);
}
for(vector<vector<int>>::iterator it1=tupleTable.begin(); it1!=tupleTable.end(); it1++) {
for(set<int>::iterator it2=allOfType.begin(); it2!=allOfType.end(); it2++) {
vector<int> newTuple;
newTuple.insert(newTuple.begin(), it1->begin(), it1->end());
newTuple.push_back(*it2);
newTable.push_back(newTuple);
}
}
synIndexMap.insert(make_pair(selectedSyn.at(i), synIndexMap.size()));
tupleTable = newTable;
}
else {
//// cout << "Syn has already been processed." << endl;
;}
}
//// cout << "End handling unprocessed selected synonyms" << endl;
if(tupleTable.size()==0)
return;
//// cout << "Begin creating tuple of answers" << endl;
for(int i=0; i<int(tupleTable.size()); i++) {
//// cout << "Creating tuple of answer " << i << endl;
string ans = "";
for(int j=0; j<int(selectedSyn.size()); j++) {
//// cout << "Adding \"" << selectedSyn.at(j) << " \"to tuple." << endl;
unordered_map<string, TypeTable::SynType>::iterator it = synTable.find(selectedSyn.at(j));
int index = synIndexMap.at(selectedSyn.at(j));
//// cout << selectedSyn.at(j) << " has index " << index << endl;
int tempAns = tupleTable.at(i).at(index);
if(it->second == TypeTable::VARIABLE) {
VarTable* varTable = pkb->getVarTable();
ans = ans + varTable->getVarName(tempAns) + " ";
//// cout << "Adding " << varTable->getVarName(tempAns) << " to tuple" << endl;
}
/*else if(it->second == TypeTable::CONSTANT) {
ConstTable* constTable = pkb->getConstTable();
ans = ans + constTable->getConst(tempAns) + " ";
//// cout << "Adding " << constTable->getConst(tempAns) << " to tuple" << endl;
}*/
else if(it->second == TypeTable::PROCEDURE && !parsedQuery.getSelectedSynIsCallProcedure().at(j)) {
ProcTable* procTable = pkb->getProcTable();
ans = ans + procTable->getProcName(tempAns) + " ";
//// cout << "Adding " << procTable->getProcName(tempAns) << " to tuple" << endl;
}
else {
ans = ans + to_string(static_cast<long long>(tempAns)) + " ";
//// cout << "Adding " << to_string(static_cast<long long>(tempAns)) << " to tuple" << endl;
}
//// cout << endl;
}
ans = ans.substr(0, ans.length()-1);
//// cout << "Final Tuple: " << ans << endl;
//// cout << endl;
//// cout << endl;
if(ansSet.find(ans) == ansSet.end()) {
result.push_back(ans);
ansSet.insert(ans);
}
}
//// cout << "End creating tuple of answers" << endl;
//// cout <<"End projecting results"<< endl;
}
std::unique_ptr<QueryTree> unpackFromQuery(const Query& query)
{
QueryParser parser;
return parser.parse(query);
}
int
query(int argc, char** argv) {
// parse arguments
parseArguments(argc, argv);
string backend = options['t'];
string indexdir = options['d'];
// check arguments: indexdir
if (indexdir.length() == 0) {
pe("Provide the directory with the index.\n");
return usage(argc, argv);
}
// check arguments: dirs
if (arguments.size() == 0) {
pe("'%s' '%s'\n", backend.c_str(), indexdir.c_str());
pe("Provide one or more files to search.\n");
return usage(argc, argv);
}
// create an index manager
IndexManager* manager = getIndexManager(backend, indexdir);
if (manager == 0) {
return usage(argc, argv);
}
IndexReader* reader = manager->indexReader();
QueryParser parser;
for (vector<string>::iterator iter = arguments.begin();
iter != arguments.end(); ++iter) {
int results = 0;
Query query = parser.buildQuery(*iter);
const uint batchsize = 10;
vector<IndexedDocument> matches = reader->query(query, 0, batchsize);
if (matches.size() != 0) {
printf ("Results for search \"%s\"\n", (*iter).c_str());
} else {
printf ("No results for search \"%s\"\n", (*iter).c_str());
}
while (matches.size() > 0) {
for (vector<IndexedDocument>::iterator it = matches.begin();
it != matches.end(); it++) {
printf ("\"%s\" matched\n", it->uri.c_str());
printIndexedDocument(*it);
}
results += (int)matches.size();
if (matches.size() == batchsize) {
// maybe there're other results
matches = reader->query(query, results + 1, 10);
} else {
matches.clear(); // force while() exit
}
}
if (results != 0) {
printf ("Query \"%s\" returned %i results\n", (*iter).c_str(),
results);
}
}
IndexPluginLoader::deleteIndexManager(manager);
return 0;
}