Re: faster m4_divert*,m4_cond

[Eric Blake]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

According to Ralf Wildenhues on 8/11/2008 9:34 PM:
|> +m4_define([_m4_cond],
|> +[m4_if(($1), [($2)], [$3],
|> +       [$#], [3], [],
|> +       [$#], [4], [$4],
|> +       [$0(m4_shift3($@))])])
|
| Sorry, but I fail to understand this.  Why does the first line of the
| m4_if put its arguments in parentheses, and why is the first argument
| not bracket-quoted?  The added test doesn't seem to expose this need.

The first argument is NOT quoted because the documented semantics of
m4_cond is that each test argument is expanded once it is encountered -
the only way to do short-circuiting is to require the caller to overquote
all tests, then have m4_cond use the lack of quotes as the trigger to
actually run the tests in sequence.  The added parenthesis are then
necessary for ensuring that m4_if sees only one leading argument, even if
the test expands into something with unquoted commas.  Yes, I should add a
test case along the lines of the following:

In autoconf 2.62, this gives a spurious warning and wrong result:
m4_cond([a,a], [a,b], [yes], [no])
m4:stdin:3: Warning: m4_if: extra arguments ignored: 5 > 4
a,b

the fixed version gives the desired result:
m4_cond([a,b], [a,b], [yes], [no])
yes

- --
Don't work too hard, make some time for fun as well!

Eric Blake             address@hidden
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkihCe8ACgkQ84KuGfSFAYAbWgCgnn6sWLRJdMGgjxKbMCza2/Cx
nWQAn0bvYG3i6mc/Q8p7BBu3fAKo7uk0
=7bbT
-----END PGP SIGNATURE-----