mn - get magic numbers from a group of files

mn [-l num] [-t desc] files

mn takes a representative selection of similarly typed files and creates
an entry suitable for putting into an /etc/magic file.  All files mentioned
on the command line are assumed to be of the same type.  The -l option
defines the maximum number of bytes into each file to consider; the default
is 1000 bytes, as this is used by most versions of file(1).  It is recommended
that this option be used to speed performance - 20 should be high enough for
most cases.

The -t option defines the description to be used. It is used verbatim, so don't
forget to quote spaces, or put the whole thing in double-quotes.  If the option
is not present, the description is read from stdin.

mn is not authorative; for instance, it can't use masks, or sense that
certain bytes are version numbers.  All tests use the = operator; for
backward compatibility, this is omitted.  The output it generates is meant to
be a starting-point - human intervention is usually required.

Note: some definitions of the isgraph macro means the exact opposite of
others (wierd!).  If you're getting garbage, change line 180 of mn.c
to include ... || isgraph(szFirstString[cCurFileDepth]) || ... in the condition.
