Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!purdue!gatech!prism!loligo!pepke
From: pepke@loligo.cc.fsu.edu (Eric Pepke)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Can anyone answer these Questions?
Message-ID: <255@loligo.cc.fsu.edu>
Date: 16 Aug 89 18:19:28 GMT
References: <840@eutrc3.urc.tue.nl> <13547@polyslo.CalPoly.EDU>
Reply-To: pepke@loligo.UUCP (Eric Pepke)
Organization: Supercomputer Computations Research Institute
Lines: 22

In article <13547@polyslo.CalPoly.EDU> dorourke@polyslo.CalPoly.EDU (David M. O'Rourke) writes:
>rcbaem@eutrc3.urc.tue.nl (Ernst Mulder) writes:
>>6) [asks about detecting bad #'s using SANE....]
>
>  Well, I know this in in-effiecient but I conver the number from str->binary
>  and then from binary->str, and see if the two str's are equal, if they are
>  then you got a #, if not you don't.  Better methods anyone??

You're half right.  01 would fail your test but would still be a perfectly
valid number.

I usually parse the string.  However, if you do that, it is almost trivial to
assemble the number as you're parsing it, so I tend to avoid SANE's conversions
altogether.

Eric Pepke                                     INTERNET: pepke@gw.scri.fsu.edu
Supercomputer Computations Research Institute  MFENET:   pepke@fsu
Florida State University                       SPAN:     scri::pepke
Tallahassee, FL 32306-4052                     BITNET:   pepke@fsu

Disclaimer: My employers seldom even LISTEN to my opinions.
Meta-disclaimer: Any society that needs disclaimers has too many lawyers.