=head1 NAME

 basecvt - radix conversion for arbitrary precision integers

=head1 SYNOPSIS

 basecvt <ibase> <obase> [values]

=head1 DESCRIPTION

The B<basecvt> program is a command-line tool for converting integers
of arbitrary precision from one radix to another.  The current version
supports radix values from 2 (binary) to 64, inclusive.  The first two
command line arguments specify the input and output radix, in base 10.
Any further arguments are taken to be integers notated in the input
radix, and these are converted to the output radix.  The output is
written, one integer per line, to standard output.

When reading integers, only digits considered "valid" for the input
radix are considered.  Processing of an integer terminates when an
invalid input digit is encountered.  So, for example, if you set the
input radix to 10 and enter '10ACF', B<basecvt> would assume that you
had entered '10' and ignore the rest of the string.

If no values are provided, no output is written, but the program
simply terminates with a zero exit status.  Error diagnostics are
written to standard error in the event of out-of-range radix
specifications.  Regardless of the actual values of the input and
output radix, the radix arguments are taken to be in base 10 (decimal)
notation.

=head1 DIGITS

For radices from 2-10, standard ASCII decimal digits 0-9 are used for
both input and output.  For radices from 11-36, the ASCII letters A-Z
are also included, following the convention used in hexadecimal.  In
this range, input is accepted in either upper or lower case, although
on output only lower-case letters are used.

For radices from 37-62, the output includes both upper- and lower-case
ASCII letters, and case matters.  In this range, case is distinguished
both for input and for output values.

For radices 63 and 64, the characters '+' (plus) and '/' (forward
solidus) are also used.  These are derived from the MIME base64
encoding scheme.  The overall encoding is not the same as base64,
because the ASCII digits are used for the bottom of the range, and the
letters are shifted upward; however, the output will consist of the
same character set.

This input and output behaviour is inherited from the MPI library used
by B<basecvt>, and so is not configurable at runtime.

=head1 SEE ALSO

 dec2hex(1), hex2dec(1)

=head1 AUTHOR

 Michael J. Fromberger <sting@linguist.dartmouth.edu>
 Thayer School of Engineering, Hanover, New Hampshire, USA

 $Date$