Re: very basic question incr

On 01/11/2012 09:36 AM, jfw wrote:
> 
> I've been programming for about 45 years.  I've been programming about 2 days
> with M4.
> 
> I want to increment or double a "variable" every time I call a macro.
> A paper "Exploiting the m4 Macro Language" by Kenneth J. Turner available on
> the web suggests:
> 
> 
> define(‘Count’,‘incr(Count)’)

That looks like the paper you are reading has used a different
typography than what m4 expects for its default quoting characters.  Are
you sure you are using backtick/apostrophe (ASCII values 0x60/0x39) and
not the Unicode single-quote characters in your actual script?  Or are
you using changequote() to select alternate quotes?

But given your description of your stack overflow, I assume you are
using the correct quoting in your script.  Rather, your problem is one
of usage.
> 
> then invoking it with
> Count

It sounds like you _wanted_ to re-define the variable Count to be the
value obtained by incrementing the previous contents of Count.  But that
is written:

define(`Count', incr(Count))

Notice that the second argument is _not_ quoted - I want incr(Count) to
be expanded and its _result_ used as an argument to define().  But in
what you wrote, you ended up defining Count to the literal string
"incr(Count)" rather than a numeric value; therefore, when you later
typed Count to cause it to be macro expanded, you triggered an infinite
expansion: Count expands to incr(Count) expands to incr(incr(Count)),
etc., until you have stack overflow.

> 
> produces m4: stack overflow
> 
> using:
> 
> define(‘Count’,incr(`Count'))
> Count produces m4: non-numeric to builtin`incr'

Now you are really worrying me - you've mixed two different quoting
styles in the same define() line, which makes me suspect that you really
aren't using the right bytes consistently.  But the warning is correct:
here, you are attempting to call incr with a single argument of the
literal string "Count", when you really meant to be calling it with a
single argument which is the number produced by the expansion of the
pre-existing macro Count.  That is, you want incr(Count) and not
incr(`Count').

> 
> using:
> define(`count',0)
> define(`count',`incr(count)')
> count
> also produces stack overflow
> 
> I would appreciate being told the simple idiom I am trying to write.  Once I
> understand incr() I assume I could use eval(x *2) to produce doubling

Indeed:
define(`count', `0')
define(`count', expr(count * 2))

-- 
Eric Blake   [email protected]    +1-919-301-3266
Libvirt virtualization library http://libvirt.org

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Public key at http://people.redhat.com/eblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBCAAGBQJPDb+eAAoJEKeha0olJ0NqLdkH/0IoPXjCl+bjaFRHwDsH9IGo
EIe8bgWL5edDSvkAzRn1zvsaUchF5p7QJIapbL+YifwoIsAQrEJ/MSI1U7sAHzXC
dnSAX4whj1LREnCPjw6gEF2f7Uks9VI7/7UzlmEf3V9H4YUds8MA/gjbbuzmJizl
om90i9UZGZUmlT44SnOXIQp9onjm7qRQmsUMrjB0lY/3wjwRhEXnWAMa0hNkbsaP
ONKSjA1C/5eZQJBbgInTl4paonobGdr5pF4D4d1uetgcNJpyS4BD717EFv2Ig5kA
OhVFvRWrfpmDGgn/hBDXRK2Bw8rAXWaofOXnNXbMxtsaWU3h3CiF6mku+cn0d2A=
=hBqr
-----END PGP SIGNATURE-----