[Octave-bug-tracker] [bug #61300] integer range might exceed upper limit

[Arun Giridhar]

Follow-up Comment #16, bug #61300 (project octave):

OK, using value() works for ascending ranges but fails for descending ranges,
giving an empty range instead. How to fix this behavior?

Updated code inside Range.h:

    void init (void)
    {
      uint64_t numer = m_limit.value () - m_base.value () + m_increment.value
();
      uint64_t denom = m_increment.value ();

      m_numel = ((m_increment == T (0)
                  || (m_limit > m_base && m_increment < T (0))
                  || (m_limit < m_base && m_increment > T (0)))
                 ? 0
                 : (numer / denom));

      m_final = m_base + (m_numel - 1) * m_increment;
    }


Wrong result for descending ranges:

>> int64 (100) : int64 (-6) : 0
ans = [](1x0)



    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?61300>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/