autoText/cpp/TextProdChar.cpp

#include <iostream>
#include <string>
#include <cstring>
#include <map>
#include <ctype.h>
#include <boost/algorithm/string.hpp>

#include "Context.hpp"
#include "LexFreq.hpp"
#include "TextProdChar.hpp"

using namespace std;

// ========================================
ostream&
operator << (ostream& os, const TextProdChar &ia) {
  string empty;
  os << ia.prevFreq.at(empty);
  os << " size = " << ia.prevFreq.size () << endl;
  for (const auto& [key, value]: ia.prevFreq)
    os << key << " = " << value.getSize () << endl;
  return os;
}


void
TextProdChar::learn (istream &in) {
  string comp;
  for (string line; getline (in, line); ) {
    for (string::iterator it (line.begin ());
	 it != line.end ();
	 ++it) {
      char c = *it;
      comp.clear ();
      comp.push_back (c);

      if (c & 0b10000000) {
	if ((c & 0b11100000) == 0b11000000) {
	  ++it; comp.push_back (*it);
	} else if ((c & 0b11110000) == 0b11100000) {
	  ++it; comp.push_back (*it);
	  ++it; comp.push_back (*it);
	} else if ((c & 0b11111000) == 0b11110000) {
	  ++it; comp.push_back (*it);
	  ++it; comp.push_back (*it);
	  ++it; comp.push_back (*it);
	} else {
	  cout << "*";
	}
      } else {
	// ASCII
	if (! Context::validChar (c))
	  continue;
      }

      if (context.empty () && (context.isSep (comp) || context.isEnd (comp)))
	continue;
      prevFreq[context.getState ()].addChar (comp);
      context.forward (comp);
    }
  }
}

void
TextProdChar::prod (ostream &out, const long &size) {
  context.reset ();
  bool nl (false), end (false);
  for (long i (0), col (0); ; ++i, ++col) {
    if (i > size)
      end = true;
    if (col > 80)
      nl = true;
    bool maj (context.empty ());
    string next (prevFreq[context.getState ()].getChar ());
    context.forward (next);
    if (maj)
      boost::to_upper (next);
    out << next;
    if (end && ! Context::getPrevSize ())
      break;
    if (context.isEnd (next) ||
	(nl && context.isSep (next))) {
      out << endl;
      nl = false;
      col = 0;
      if (end && context.isEnd (next))
	break;
    }
  }
  out << endl;
}

// ========================================