deprecating allocelems (replaced by numElements) and maxlength (replaced by
bufferSize); avoid increasing memory allocation for strings when using variable
width character sets and increasing the number of elements in a variable during
executemany(). Thanks to Don Reid for pointing out this issue.
support within Python 2.x; move away from character semantics which Oracle is
deprecating anyway to byte semantics which should hopefully eliminate the
problem with a backend character set of UTF-8.