BoyerMooreHorspool.h

Go to the documentation of this file.

 
#pragma once
 
#include <alib/set>
#include <alib/map>
 
#include <common/ranked_symbol.hpp>
 
#include <tree/properties/BadCharacterShiftTable.h>
#include <tree/properties/SubtreeJumpTable.h>
#include <tree/properties/ExactSubtreeRepeatsNaive.h>
#include <tree/exact/BackwardOccurrenceTest.h>
 
#include <tree/ranked/PrefixRankedBarTree.h>
#include <tree/ranked/PrefixRankedBarPattern.h>
#include <tree/ranked/PrefixRankedBarNonlinearPattern.h>
 
namespace arbology {
 
namespace exact {
 
class BoyerMooreHorspool {
public:
    template < class SymbolType >
    static ext::set < unsigned > match ( const tree::PrefixRankedBarTree < SymbolType > & subject, const tree::PrefixRankedBarTree < SymbolType > & pattern );
    template < class SymbolType >
    static ext::set < unsigned > match ( const tree::PrefixRankedBarTree < SymbolType > & subject, const tree::PrefixRankedBarPattern < SymbolType > & pattern );
    template < class SymbolType >
    static ext::set < unsigned > match ( const tree::PrefixRankedBarTree < SymbolType > & subject, const tree::PrefixRankedBarNonlinearPattern < SymbolType > & pattern );
 
};
 
template < class SymbolType >
ext::set < unsigned > BoyerMooreHorspool::match ( const tree::PrefixRankedBarTree < SymbolType > & subject, const tree::PrefixRankedBarTree < SymbolType > & pattern ) {
    return match ( subject, tree::PrefixRankedBarPattern < SymbolType > ( pattern ) );
}
 
template < class SymbolType >
ext::set < unsigned > BoyerMooreHorspool::match ( const tree::PrefixRankedBarTree < SymbolType > & subject, const tree::PrefixRankedBarPattern < SymbolType > & pattern ) {
    ext::set < unsigned > occ;
    ext::map < common::ranked_symbol < SymbolType >, size_t > bcs = tree::properties::BadCharacterShiftTable::bcs ( pattern ); //NOTE: the subjects alphabet must be a subset or equal to the pattern
    ext::vector < int > subjectSubtreeJumpTable = tree::properties::SubtreeJumpTable::compute ( subject );
 
     // index to the subject
    unsigned i = pattern.getContent ( ).size ( ) - 1;
 
     // main loop of the algorithm over all possible indexes where the pattern can start
    while ( i < subject.getContent ( ).size ( ) ) {
        // pair j and offset
        ext::pair < int, int > jOffset = tree::exact::BackwardOccurrenceTest::occurrence ( subject, subjectSubtreeJumpTable, pattern, i );
 
         // match was found
        if ( jOffset.first == -1 ) occ.insert ( jOffset.second + 1);
 
         // shift heuristics
        i += bcs[subject.getContent ( )[i]];
    }
 
    return occ;
}
 
template < class SymbolType >
ext::set < unsigned > BoyerMooreHorspool::match ( const tree::PrefixRankedBarTree < SymbolType > & subject, const tree::PrefixRankedBarNonlinearPattern < SymbolType > & pattern ) {
    ext::set < unsigned > occ;
    ext::map < common::ranked_symbol < SymbolType >, size_t > bcs = tree::properties::BadCharacterShiftTable::bcs ( pattern ); //NOTE: the subjects alphabet must be a subset or equal to the pattern
 
    ext::vector < int > subjectSubtreeJumpTable = tree::properties::SubtreeJumpTable::compute ( subject );
    tree::PrefixRankedBarTree < unsigned > repeats = tree::properties::ExactSubtreeRepeatsNaive::repeats ( subject );
 
     // index to the subject
    unsigned i = pattern.getContent ( ).size ( ) - 1;
 
     // main loop of the algorithm over all possible indexes where the pattern can start
    while ( i < subject.getContent ( ).size ( ) ) {
        // pair j and offset
        ext::pair < int, int > jOffset = tree::exact::BackwardOccurrenceTest::occurrence ( subject, subjectSubtreeJumpTable, repeats, pattern, i );
 
         // match was found
        if ( jOffset.first == -1 ) occ.insert ( jOffset.second + 1);
 
         // shift heuristics
        i += bcs[subject.getContent ( )[i]];
    }
 
    return occ;
}
 
} /* namespace exact */
 
} /* namespace arbology */