buildEsa.cpp

buildEsa.cpp

#include <iostream>
#include <string>
#include <vector>
#include <map>
#include "esa.hxx"

using namespace std;

int readFile(const char* fn, vector<int>& T){
  FILE* fp = fopen(fn, "rb");
  if (fp == NULL){
    cerr << "cannot open " << fn << endl;
    return -1;
  }

  if (fseek(fp, 0, SEEK_END) != 0){
    cerr << "cannot fseek " << fn << endl;
    fclose(fp);
    return -1;
  }
  int n = ftell(fp);
  rewind(fp);
  if (n < 0){
    cerr << "cannot ftell " << fn << endl;
    fclose(fp);
    return -1;
  }
  T.resize(n);
  if (fread(&T[0], sizeof(unsigned char), (size_t)n, fp) != (size_t) n){
    cerr << "fread error " << fn << endl;
    fclose(fp);
    return -1;
  }

  fclose(fp);
  return 0;
}

void printSubstr(const vector<int>& T, const int beg){
  for (int i = beg; i < T.size(); ++i){
    int c = T[i];
    cout << (isspace((char)c) ? '_' : (char)c);
  }
}

int main(int argc, char* argv[]){

  vector<int> T;

  istreambuf_iterator<char> isit(cin);
  istreambuf_iterator<char> end;

  size_t origLen = 0;
  while (isit != end){
    T.push_back((unsigned char)(*isit++));
  }
  origLen = T.size();

  vector<int> SA(T.size());
  vector<int> L (T.size());
  vector<int> R (T.size());
  vector<int> D (T.size());

  int k = 0x100;

  cerr << "    n:" << T.size() << endl;
  cerr << "alpha:" << k        << endl;

  int nodeNum = 0;
  if (esaxx(T.begin(), SA.begin(), 
	    L.begin(), R.begin(), D.begin(), 
	    (int)T.size(), k, nodeNum) == -1){
    return -1;
  }

  cout << " Suffix Array:" << endl; 
  cout << "i\tSA\tL\tR\tD\tsubstring:" << endl;
  cout << "--------------------------------" << endl;
  for (int i = 0; i < T.size(); i++) {
    cout << i << "\t" << SA[i] << "\t" << L[i] << "\t" << R[i] << "\t" << D[i] << "\t";
    printSubstr(T, SA[i]);
    cout << endl;
  }
  cout << endl;

  return 0;
}